JP4007242B2 - Semiconductor device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a bipolar self-extinguishing semiconductor device that operates in a bipolar mode, such as a bipolar transistor or IGBT (insulated gate bipolar transistor) used in a power converter.
[0002]
[Prior art]
In a MOSFET used for a gate array, a basic cell for flowing a relatively small current for signals and a basic cell for flowing a large current used for an output unit are usually formed separately. Therefore, the usage rate of basic cells in the gate array and the degree of freedom in circuit design are reduced. In order to solve this problem, in the MOSFET constituting the basic cell, the gate electrode is divided into a plurality of parts, and the voltage is selectively applied to the individual electrodes, so that the channel length is changed to have a plurality of types of operation states. By doing so, it is known that the usage rate of basic cells in the gate array is improved and the degree of freedom in circuit design is secured (see Patent Document 1).
[0003]
In a MOSFET with a built-in overcurrent protection function, the gate electrode of the MOSFET to be detected is divided into several parts in the gate width direction (perpendicular to the channel length), and the potential of the divided gate electrode is lowered below the threshold value. It is known that a drive element is provided and the drain current is narrowed down and overcurrent protection is performed by making the potential of the gate electrode sequentially divided at the overcurrent level equal to or less than a threshold value (see Patent Document 2).
Further, a plurality of semiconductor elements having MOS structure gates are formed on a chip, and a gate signal with a time difference is given to the gates formed in each element to gradually increase / decrease the ON region, collector current at turn-on / turn-off, Relaxing the rate of change of collector-emitter voltage over time, peaks and vibrations exceeding the rating generated at the rise of the current waveform at turn-on, peaks and vibrations exceeding the rating generated at the rise of the voltage waveform at turn-off Is known to prevent malfunction and destruction of the semiconductor device (see Patent Document 3).
[0004]
In recent years, power semiconductor devices have been increased in capacity, and devices having a current capacity of about several hundreds of A to several thousand A have been announced. In general, one power semiconductor device has a structure in which a plurality of basic cells, which are the basic structure of a power semiconductor device, are repeatedly arranged in an island shape or a plurality of stripe shapes. This can be done by increasing the size or connecting multiple chips in parallel.
Power semiconductor devices applied to power converters are roughly classified into non-self-extinguishing devices such as thyristors that do not have turn-off capability and self-extinguishing devices such as transistors that have turn-off capability. Since a non-self-extinguishing type device requires a commutation circuit to cut off the main current and the operating frequency is low, a self-extinguishing type device is usually used in a power converter such as an inverter circuit. . Typical bipolar devices in this self-extinguishing device include bipolar transistors, IGBTs (insulated gate bipolar transistors), GTO thyristors (gate turn-off thyristors), and MOS thyristors. A unipolar device is a MOSFET.
[0005]
In a bipolar self-extinguishing semiconductor device, the on state and the off state are realized by increasing or decreasing the amount of carriers in the semiconductor by applying a current or voltage to the control electrode. When the amount of carriers inside the semiconductor is large, the semiconductor device is turned on, and when it is small (almost no), the semiconductor device is turned off. The amount of carriers is the sum of the donor or acceptor concentration of the semiconductor substrate and the amount of accumulated carriers.
A period during which the ON state transitions to the OFF state is a turn-off time, and a period during which excess carriers accumulated in the semiconductor are swept out (a period during which the carriers are discharged).
When a bipolar self-extinguishing semiconductor device is turned off, it takes time to discharge and disappear excess carriers accumulated in the semiconductor, so that there is a time lag between the turn-off signal input to the control electrode and the turn-off completion. This time lag is the turn-off time.
[0006]
The amount of excess carriers stored in this bipolar self-extinguishing semiconductor device is less dependent on the energizing current when the conduction area is constant, and is smaller when a small current is applied than when a large current is applied. The amount of excess carriers to be accumulated does not decrease with the ratio of the energization current, but is a relatively large value. On the other hand, the amount of excess carriers accumulated is proportional to the current-carrying area (the area of the active region that is operating).
When a bipolar self-extinguishing semiconductor device is operated in an inductive load state and in a small current state, as described above, a relatively large number of accumulated carriers exist in the semiconductor, and this excess carrier is small through the inductive load. Since the current is gradually discharged to the outside, it takes time to discharge excess carriers, and the turn-off time becomes long.
[0007]
FIG. 5 is a plan view of a main part of a conventional IGBT chip. FIG. 5A shows a case where the active region is not divided, and FIG. 5B shows a case where the active region is divided. The sizes of the IGBT chips 51 and 61 are both 21.5 mm □.
In FIG. 2A, a gate electrode (not shown) is formed on the active region 42 with the same size as the active region 42 and is connected to the gate pad 53 through a connection portion 54.
In FIG. 6B, the active region 66 is divided into 16 unit active regions 62, and the size of each unit active region 62 is about 2 mm × about 8 mm. Further, a gate electrode (not shown) is divided into 16 pieces corresponding to the unit active region 62, and the gate electrode (not shown) formed on each unit active region 62 is connected to the gate runner 64 at the contact portion 63. The gate runner 64 is connected to the gate pad 65.
[0008]
6A and 6B are diagrams showing a turn-off waveform in the IGBT in an inductive load state. FIG. 6A shows a case where the collector current is 125 A, and FIG. 6B shows a case where the collector current is 6.2 A. This is a turn-off waveform of a 4.5 kV class IGBT containing the IGBT chip 61 of FIG. In the figure, Vbus is an ultimate voltage, Rg is a gate resistance connected to the outside, Tj is a junction temperature, Ic is a collector current, VCE is a collector voltage, and VGE is a gate voltage.
When the collector current is 125 A, the turn-off time is 2.5 μs, whereas when the collector current is 6.2 A, the turn-off time is significantly increased to 6 μs. This turn-off waveform is the same for the chip of FIG. Further, the factors that increase the turn-off time at a small current are as described above.
[0009]
[Patent Document 1]
JP-A-9-260660 [Patent Document 2]
Japanese Patent No. 3126540 [Patent Document 3]
Japanese Patent Laid-Open No. 8-32064
[Problems to be solved by the invention]
In a power conversion device such as an inverter, a dead time is provided in order to prevent the upper and lower arms from being short-circuited. This dead time is the period from when the upper arm device is turned off to when the lower arm device is turned on (or the period from when the lower arm device is turned off to when the upper arm device is turned on). Yes, it is set by adding a slight margin to the turn-off time. Therefore, when the turn-off time becomes longer, the ratio of the rest period to the conduction period of the device increases. In particular, this ratio increases in high-frequency operation, which is an obstacle to increasing the frequency of the inverter output and obtaining a sine wave with less harmonic components.
[0011]
An object of the present invention is to provide a bipolar self-extinguishing semiconductor device capable of solving the above-described problems and shortening the turn-off time even in an inductive load state and in a small current energizing state.
[0012]
[Means for Solving the Problems]
To achieve the above object, a plurality of active regions having different sizes formed on a semiconductor substrate, a first main electrode formed collectively on a first main surface of the plurality of active regions, and each active region A control terminal formed in each region and a second main electrode formed on the second main surface of the plurality of active regions, and a large active region when the main current at the time of interruption is large The large active region is operated by the control signal sent to the control terminal of the main current, and when the main current is small, the small active region is operated by the control signal sent to the control terminal of the small active region. The size of the active region that operates according to the size is changed.
[0013]
A plurality of active regions formed on the semiconductor substrate; a first main electrode formed collectively on a first main surface of the plurality of active regions; a control terminal formed on each active region; And a second main electrode formed on the second main surface of the plurality of active regions, and when the main current at the time of interruption is large, each control signal sent to the control terminal of each active region When the active region is operated simultaneously and the main current is small, the active region is operated according to the magnitude of the main current by operating the active region with a control signal sent to the control terminal of the active region. The size of the active region to be changed is changed.
Also, two active regions formed in the semiconductor substrate, a first main electrode formed collectively on the first main surface of the two active regions, and a high gate threshold value formed in the first active region A first MOS gate structure having a voltage, a second MOS gate structure having a low gate threshold voltage formed in the second active region, and a connection between the first MOS gate structure and the second MOS gate structure And a second main electrode formed on the second main surface of the two active regions, and when the main current at the time of interruption is large, it is higher than a high threshold voltage A gate voltage is applied to the control terminal to simultaneously operate the first and second active regions, and when the main current is small, an intermediate gate voltage between a low gate threshold voltage and a high gate threshold voltage is Second active region applied to control terminal By operating, changing the size of the active region to be operated in accordance with the magnitude of the main current.
[0014]
In addition, the first to n-th n active regions formed on the semiconductor substrate, a first main electrode collectively formed on the first main surface side of the n active regions, and a first A first MOS gate structure having a first gate threshold voltage which is the maximum voltage formed in the active region; and an nth gate threshold which is a minimum voltage formed in the nth active region. An n-th MOS gate structure having a voltage, an m-th MOS gate structure having an m-th gate threshold voltage, which is an intermediate voltage formed in the m-th active region in the middle, and the first MOS gate The structure includes a common control terminal connected to the nth MOS gate structure from the structure, and a second main electrode formed on the second main surface side of the n active regions, and the main current at the time of interruption is When larger than the first gate threshold voltage When a main gate voltage is applied to the control terminal to simultaneously operate the first to nth active regions, and the main current is small, the nth gate threshold voltage and the gate threshold voltage Next, a gate voltage having a magnitude between the next n-1th gate threshold voltage is applied to the control terminal to operate the nth active region, and the main current has an intermediate magnitude. In the m-th case, a gate voltage having a magnitude between the m-th gate threshold voltage and the (m-1) -th gate threshold voltage that is higher than the gate threshold voltage is applied to the control terminal. Applying and operating the m-th to n-th active regions, changing the size of the active region that operates according to the size of the main current, and operating according to the size of the main current The size of the active region to be changed is changed.
[0015]
The semiconductor device may be a bipolar self-extinguishing semiconductor device. The bipolar self-extinguishing semiconductor device may be any one of a bipolar transistor, an insulated gate bipolar transistor, a gate turn-off thyristor, and a MOS thyristor.
Also, in a semiconductor device in which a plurality of bipolar self-extinguishing semiconductor chips are connected in parallel and housed in the same package, a separate control signal is sent to the control terminal formed on each chip, and the main current at the time of interruption When the main current is large, the number of chips to be operated is increased, and when the main current is small, the number of chips to be operated is decreased.
[0016]
The bipolar self-extinguishing semiconductor chip may be any one of a bipolar transistor chip, an insulated gate bipolar transistor chip, a gate turn-off thyristor chip, and a MOS thyristor chip.
[Action]
The active area of the device is divided, and the divided active areas are individually controlled. When the main current at the time of turn-off is small, if the area of the active region to be operated is reduced, the amount of accumulated carriers is reduced, and the amount of carriers that need to be swept out at the time of turn-off can be reduced. As a result, the turn-off time can be shortened even when the main current is small.
[0017]
DETAILED DESCRIPTION OF THE INVENTION
[Example 1]
FIG. 1 is a plan view of an essential part of a semiconductor device according to a first embodiment of the present invention. This figure is a plan view of an IGBT chip. When a plurality of chips are stored in a flat package, a pressure contact type flat IGBT is obtained.
The size of the IGBT chip 1 is 21.5 mm □, the unit active region 2 is composed of a large number of cells, and the structure of each cell is not shown, but the well-known unit emitter region, unit gate electrode, unit It consists of a well region, a unit drift region and a unit collector region. That is, the cell is a region obtained by dividing the IGBT in the direction of the semiconductor substrate by one well region (unit well region). A large number of these cells aggregate to form one unit active region. Further, one or a plurality of unit active regions become active regions that operate. An emitter electrode is formed on the emitter region of the active region, and a collector electrode is formed on the collector region.
[0018]
In FIG. 1, the active region is composed of 16 unit active regions 2, and the size of each unit active region 2 is about 2 mm × 8 mm. 15 unit active regions 2 constitute a large active region 7, and one unit active region 2 constitutes a small active region 8. A polysilicon gate electrode (not shown) formed on each of the 15 unit active regions 2 is connected to an aluminum gate runner 4 at a connection portion 3. The gate runner 4 is connected to an aluminum gate pad 5 (control terminal). Connecting. Further, a gate electrode (not shown) formed on the remaining one unit active region 2 is connected to an aluminum gate pad 6 (control terminal) at the connection portion 3, and the gate pad 6 is electrically connected to the gate pad 5. Insulated. A common emitter electrode (not shown) is formed of aluminum on the large active region 7 and the small active region 8.
[0019]
2A and 2B are diagrams showing an IGBT turn-off waveform in an inductive load state. FIG. 2A shows a case where the collector current is 125 A, and FIG. 2B shows a case where the collector current is 6.2 A. This is a turn-off waveform of a 4.5 kV class IGBT containing the IGBT chip 1 of FIG. As described above, Vbus in the figure is an ultimate voltage, Rg is a gate resistance connected to the outside, Tj is a junction temperature, Ic is a collector current, VCE is a collector voltage, and VGE is a gate voltage.
When the collector current is 6.2 A, the gate voltage is applied only to the gate pad 6 to operate the small active region 8 (one unit active region 2). When the collector current is 125 A, the gate voltage is applied to both the gate pad 5 and the gate pad 6 at the same time, and all 16 unit active regions 2 (the large active region 7 and the small active region 8) are operated simultaneously.
[0020]
When the collector current is 125 A, the active regions 7 and 8 that operate are the 16 unit active regions 2 and the area thereof is the same as that of the conventional IGBT. The turn-off waveforms are the same. Therefore, the turn-off time is the same as that of the conventional IGBT, which is 2.5 μs. When only the gate pad 5 is operated, the large active region 7 (15 unit active regions 2) operates, and the turn-off time is almost the same because the area of the active region to operate is almost the same as 16. Absent. Of course, even if the 15 unit active regions 2 are operated by applying a gate voltage only to the gate pad 5, the turn-off time is hardly changed because the area of the active region to be operated is only reduced by 1/16.
[0021]
On the other hand, in the case of turn-off of 6.2 A, a gate voltage is applied only to the gate pad 6 to operate a small active region 8 (one unit active region 2). The amount of carriers to be issued is significantly smaller than that of the conventional IGBT. Therefore, the turn-off time is reduced to 3.5 μs, which is about half that of the conventional IGBT.
As described above, when the collector current is small, the area of the active region to be operated is reduced, and the amount of carriers swept out at the time of turn-off is reduced, so that the turn-off time can be shortened.
[0022]
As described above, the same effect can be obtained in the case of bipolar self-extinguishing semiconductor devices such as bipolar transistors, GTO thyristors, and MOS (voltage driven) thyristors. In this example, the ratio of the two large and small active regions is 16: 1 in terms of the number of unit active regions. However, the area of the small active region may be appropriately changed according to the required current capacity. The number used may be changed.
[Example 2]
FIG. 3 is a plan view of an essential part of a semiconductor device according to the second embodiment of the present invention. This is a case where the present invention is applied to an IGBT module 10 in which a plurality of IGBT chips 11 are connected in parallel.
[0023]
The IGBT chip 11 is fixed on the copper pattern 19 formed on the insulating substrate 18, and the diode chip 14 is fixed on the copper pattern 20. The copper pattern 19 and the cathode electrode 15 of the diode chip 14 are connected by a boarding wire, and the copper pattern 20 and the emitter electrode 12 of the IGBT chip 11 are connected by a bonding wire 16. The control external lead terminal 22 insulated by the copper pattern 19 and the insulating film 21 is connected to the gate pad 13 of the three IGBT chips 11 and the bonding wire 17, and the control external lead terminal 23, the single IGBT chip 11 and the bonding wire are connected. 17 to connect. Thus, the IGBT module 10 is formed by housing the insulating substrate 18 to which the plurality of IGBT chips and the diode chip 14 are fixed in a resin package (not shown).
[0024]
As described above, four IGBT chips 11 having the same operation area are used, and only one IGBT chip is individually controllable. When the current to be turned off is small, the control external lead-out terminal 23 is connected. One IGBT chip 11 is operated by applying a gate voltage to the gate pad 13 of one IGBT.
In the case of a normal current with a large turn-off current, the gate voltage is applied to the gate pads 13 of the four IGBT chips 11 connected to both of the control external lead-out terminals 22 and 21, and the four IGBT chips 11 are operated. The gate voltages are applied to the gate pads 13 of the three IGBT chips 11 connected to the external lead-out terminals 22 to operate the three IGBT chips 11.
[0025]
In this way, when the collector current at the time of turn-off is small, the turn-off time can be shortened by reducing the number of IGBT chips 11 to be operated and reducing the operation area, as in the first embodiment.
Moreover, although it comprised with the IGBT chip | tip of the same area here, you may comprise with an IGBT chip | tip with a different area. Further, the IGBT chip may be housed in a flat package instead of a module to form a flat IGBT.
Of course, the same effect can be obtained by storing a bipolar self-extinguishing semiconductor chip such as a power bipolar transistor chip, a GTO thyristor chip, and a MOS thyristor chip instead of the IGBT chip.
Example 3
FIG. 4 is a plan view of the main part of the semiconductor device according to the third embodiment of the present invention. As in FIG. 1, the active region is divided into 16 unit active regions 32, the gate threshold voltage of a large active region 37 composed of 15 unit active regions 32 is increased, and one unit active region is formed. The gate threshold voltage of the small active region 38 composed of 32 is lowered.
[0026]
There are various methods for adjusting the gate threshold voltage, such as changing the thickness of the gate insulating film and changing the impurity concentration of the channel forming portion. Here, the gate threshold voltage is increased to increase the threshold value. The voltage was reduced to a lower threshold voltage.
When the IGBT chip 31 is operated, the gate voltage applied to the gate pad 35 is set between a high gate threshold voltage and a low gate threshold, so that only the small active region 38 having a low gate threshold is operated. The large active region 37 and the small active region 38 are simultaneously operated by setting the gate voltage higher than the high gate threshold voltage.
[0027]
When the collector current at turn-off is small, the former gate voltage is set, and when the collector current at turn-off is large, the latter gate voltage is set to obtain the same effect as in the first embodiment. it can.
In this embodiment, the gate threshold voltage is set to 2 levels. However, as the number of levels is further increased and the collector current is reduced, the same effect can be obtained even if the active regions that are sequentially operated are reduced. Of course. In the figure, 33 is a connecting portion, and 34 is a gate runner.
In each of the above-described embodiments, the gate signal is changed depending on the magnitude of the collector current at the time of turn-off. The gate signal may be switched so as to turn on only a predetermined small active region at the time of turn-on.
[0028]
【The invention's effect】
According to the present invention, when the collector current at turn-off is large, the area of the active region in which the semiconductor device operates is increased, and when the collector current is small, the area of the active region in operation is decreased to reduce the collector current. The turn-off time in the inductive load state can be greatly shortened.
By using this semiconductor device for power converters such as inverters, dead time in a minute load current state can be shortened in an inductive load state, and the operating frequency of the converter device can be increased and the output waveform harmonics can be reduced. Fewer sine waves can be achieved.
[Brief description of the drawings]
FIG. 1 is a plan view of a principal part of a semiconductor device according to a first embodiment of the present invention. FIG. 2 is a diagram showing a turn-off waveform of an IGBT in an inductive load state. FIG. 3 is a plan view of the main part of the semiconductor device according to the second embodiment of the present invention. FIG. 4 is a plan view of the main part of the semiconductor device according to the third embodiment of the present invention. 5A and 5B are plan views of a main part of a conventional IGBT chip, in which FIG. 6A is a diagram when the active region is not divided, and FIG. 6B is a diagram when the active region is divided. The figure which shows the turn-off waveform in IGBT, (a) when collector current is 125A, (b) when collector current is 6.2A
1, 11, 31 IGBT chip 2, 32 Unit active region 3, 33 Connection portion 4, 34 Gate runner 5, 6, 35 Gate pad 7, 37 Large active region 8, 38 Small active region 10 IGBT module 12 Emitter electrode 13 Gate Pad 14 Diode chip 15 Cathode electrode 16, 17 Bonding wire 18 Insulating substrate 19, 20 Copper pattern 21 Insulating film 22, 23 Control external lead-out terminal

Claims (8)

A plurality of active regions of different sizes formed on a semiconductor substrate; a first main electrode formed in a lump on a first main surface of the plurality of active regions; and a control terminal formed in each active region And a second main electrode formed on the second main surface of the plurality of active regions, and when the main current at the time of interruption is large, a large control signal is sent to the control terminal of the large active region When the active region is operated and the main current is small, the small active region is operated by the control signal sent to the control terminal of the small active region. A semiconductor device characterized by changing the thickness. A plurality of active regions formed in a semiconductor substrate; a first main electrode formed collectively on a first main surface of the plurality of active regions; a control terminal formed in each active region; A second main electrode formed on the second main surface of each active region, and when the main current at the time of interruption is large, each active region is simultaneously controlled by a control signal sent to the control terminal of each active region When the main current is small, the active region is operated according to the magnitude of the main current by operating a part of the active region with a control signal sent to the control terminal of the part of the active region. A semiconductor device characterized in that the size is changed. Two active regions formed in a semiconductor substrate, a first main electrode formed collectively on the first main surface of the two active regions, and a high gate threshold voltage formed in the first active region A first MOS gate structure having a low gate threshold voltage formed in the second active region, and the first MOS gate structure and the second MOS gate structure are connected to each other. A common control terminal and a second main electrode formed on the second main surface of the two active regions, and when the main current at the time of interruption is large, the threshold value of the first MOS gate structure A gate voltage higher than the voltage is applied to the control terminal to simultaneously operate the first and second active regions. When the main current is small, the second MOS is lower than the threshold voltage of the first MOS gate structure. Threshold voltage of gate structure Ri high gate voltages by operating the second active region is applied to the control terminal, the semiconductor device characterized by changing the size of the active region to be operated in accordance with the magnitude of the main current. First to n-th n active regions formed on a semiconductor substrate, a first main electrode collectively formed on the first main surface side of the n active regions, and a first active region A first MOS gate structure having a first gate threshold voltage, which is the maximum voltage formed at the same time, and an nth gate threshold voltage, which is the minimum voltage formed in the nth active region. An nth MOS gate structure having an mth gate threshold voltage, which is an intermediate voltage formed in the mth active region in the middle thereof, and the first MOS gate structure A common control terminal connected to the nth MOS gate structure portion and a second main electrode formed on the second main surface side of the n active regions, and when the main current at the time of interruption is large, A gate voltage higher than the first gate threshold voltage The first to nth active regions are simultaneously operated by applying to the control terminal, and when the main current is small, the nth gate threshold voltage and the second higher threshold voltage than the gate threshold voltage are applied. When a gate voltage having a magnitude between the (n-1) th gate threshold voltage is applied to the control terminal to operate the nth active region, and the main current is the middle magnitude and the mth Applies a gate voltage having a magnitude between the mth gate threshold voltage and the (m-1) th gate threshold voltage, which is higher than the gate threshold voltage, to the control terminal; By operating the mth to nth active regions, the size of the active region that operates according to the size of the main current is changed, so that the active region that operates according to the size of the main current is changed. A semiconductor device characterized in that the size is changed. The type semiconductor device according to any one of claims 1 to 4, wherein the semiconductor device is a bipolar self-extinguishing semiconductor device. 6. The semiconductor device according to claim 5, wherein the bipolar self-extinguishing semiconductor device is any one of a bipolar transistor, an insulated gate bipolar transistor, a gate turn-off thyristor, and a MOS thyristor. In a semiconductor device in which a plurality of bipolar self-extinguishing semiconductor chips are connected in parallel and housed in the same package, an individual control signal is sent to the control terminal formed on each chip, and the main current at the time of interruption is large. According to the semiconductor device, the number of chips to be operated is switched, the number of chips to be operated is increased when the main current is large, and the number of chips to be operated is decreased when the main current is small. 8. The semiconductor according to claim 7, wherein the bipolar self-extinguishing semiconductor chip is one of a bipolar transistor chip, an insulated gate bipolar transistor chip, a gate turn-off thyristor chip, and a MOS thyristor chip. apparatus.

Images